In entering upon any speculations or enquiries concerning the nature of the interior of our globe, it is necessary before all things that we should clearly realise in our minds how small and almost infinitesimal is that part of the earth's mass which can be subjected to direct examination. The distance from the surface to the centre of our globe is nearly 4,000 miles, but the deepest mines do not penetrate to much more than half a mile from the surface, and the deepest borings fall far short of a mile in depth. Sometimes, it is true, the geologist finds means for drawing inferences as to the nature of the rocks at depths of ten or fifteen miles below the surface; but the last-named depth must be regarded as the utmost limit of that portion of our globe which can be made the object of direct observation and study. This thin exterior film of the earth's mass, which the geologist is able to investigate, we call the 'crust of the globe'; but it must be remembered that in using this term, it is not intended to imply that the outer part of our globe differs in any essential respect from the interior. The term 'crust of the globe' is employed by geologists as a convenient way of referring to that portion of the earth which is accessible to their observation.

But if we are unable to make direct investigations concerning the nature of the internal portions of the globe, there are nevertheless a number of facts from which we may draw important inferences upon the subject. These facts and the inferences based upon them we shall now proceed to consider.

First in importance among these we may mention the results which have been obtained by weighing our globe. Various methods have been devised for accomplishing this important object, and the conclusions arrived at by different methods agree so closely with one another, that there is no room for doubt as to the substantial accuracy of those results. It may be taken as proved beyond the possibility of controversy that our globe is equal in weight to five and a half globes of the same size composed of water, or, in other words, that the average density of the materials composing the globe is five and a half times as great as that of water.

Now the density of the materials which compose the crust of the globe is very much less than this, varying from about two-and-one-third to three times that of water. Hence we are compelled to conclude that the interior portions of the globe are of far greater density than the exterior portions; that, as a matter of fact, the mass of the globe is composed of materials having twice the density of the rocks exposed at the surface.

It has been sometimes argued that as all materials under intense pressure appear to yield to an appreciable extent, and to allow their particles to be packed into a smaller compass, we may find in this fact an explanation of the great density of the internal parts of the globe. It has in fact been suggested that under the enormous pressure which must be exerted by masses of rock several thousand feet in thickness, the materials of which our earth is composed may be compelled to pack themselves into less than one-half the compass which they occupy at the surface. But the ascription of such almost unlimited compressibility to solid substances can be supported neither by experiment nor analogy. Various considerations point to the probability that solid bodies yield to pressure up to a certain limit and no farther, and that when this limit is reached an increase in pressure is no longer attended with a reduction in bulk.

If then we are compelled to reject the idea of the unlimited compressibility of solid substances, we must conclude that the interior portions of our globe are composed of materials of a different kind from those which occur in its crust. And this conclusion, as we shall presently see, is borne out by a number of independent facts.

The study of the materials ejected from volcanic vents proves that even at very moderate depths there exist substances differing greatly in density, as well as in chemical composition. The lightest lavas have a specific gravity of 2·3, the heaviest of over 3. And that materials of even greater density are sometimes brought by volcanic action from the earth's interior, we have now the clearest proofs.

But in considering a question of this kind, it will be well to remember that analogy may furnish us with hints upon the subject which may prove to be by no means unimportant. There is no question upon which modern science has wrought out a more complete revolution in our ideas, than that of the relation of our earth to the other bodies of the universe. We know, as the result of recent research, that our globe is one of a great family of bodies, moving through space in similar paths and in obedience to the same laws. A hundred years ago the primary and secondary planets of the solar system could be almost numbered upon the fingers; now we recognise the fact that they exist in countless millions, presenting every variety of bulk from masses 1,400 times as large as our earth down to the merest planetary dust. Between the orbits of Mars and Jupiter, more than 200 small planets have been recognised as occurring, and every year additions are made to the number of these asteroids. Comets have now been identified with streams of such planetary bodies, of minute size, moving in regular orbits through our system. The magnificent showers of 'shooting-stars' have been proved to be caused by the passage of the earth through such bands of travelling bodies, and 'the zodiacal light' finds its most probable explanation in the supposition that the sun is surrounded by a great mass of such minute planets. Every increase in the power of the telescope reveals to us the existence of new secondary planets or moons, revolving about the primaries; and the wonderful system of the Saturnian rings is now explained by the proved existence of great streams of such secondary planets circling around it. The solar system was formerly conceived of as a vast solitude through which a few gigantic bodies moved at awful distances from one another. Now we know that the supposed empty void is traversed by countless myriads of bodies of the most varied dimensions, all moving in certain definite paths, in obedience to the same laws, ever acting and reacting upon each other, and occasionally coming into collision.

There are not wanting further facts to prove that the other planets are like our own in many of their phenomena and surroundings. In some of them atmospheric phenomena have been detected, such as the formation of clouds and the deposition of snow, so that the external forces at work on our globe act upon them also. And that internal forces, like those we have been considering in the case of our earth, are at work in our neighbours, is proved by the great solar storms and the condition of the moon's surface.

But the results of spectrum-analysis in recent years have furnished new facts in proof of the close relationship of our earth to the numerous similar bodies by which it is surrounded. So far as observation has yet gone we have reason for believing that not only the members of the solar system, but the more distant bodies of the universe, are all composed of the same elementary substances as those which enter into the composition of our globe.

The most satisfactory information concerning the composition and nature of other planetary bodies is derived from the study of those small planets which occasionally come into collision with our globe, and which have their own proper motion in space thereby arrested. These meteorites, as such falling planetary bodies are called, have justly attracted great attention, and their fragments are treasured as the most valuable objects in our museums.

The first fact concerning these meteorites, which it is necessary to notice, is that they are composed of the same chemical elements as occur in the earth's crust. No element has yet been found in any meteorite which was not previously known as existing in the earth, and of the sixty-five or seventy known terrestrial elements no less than twenty-two have already been detected in meteorites.

There are, however, a dozen elements which occur in overwhelming proportions in the earth's crust. We shall probably not be going too far in saying that these twelve elements—namely, oxygen, silicon, aluminium, calcium, magnesium, sodium, potassium, iron, carbon, hydrogen, sulphur, and chlorine—make up amongst them not less than 999 out of 1,000 parts of the earth's crust, and that all the other fifty or sixty elements are 80 comparatively rare that they do not constitute when taken altogether more than one part in 1,000 of the rocks of the globe. Now all of these twelve common terrestrial elements occur in meteorites, and the fact that the rarer terrestrial elements have not as yet been found in them will not surprise anyone, who remembers how small is the bulk of all the specimens of these meteorites existing in our museums.

We have hitherto insisted on the points of resemblance in the chemical composition of meteorites and that of the rocks of the globe, but we shall now have to indicate some very important points in which they differ.

While in the rocks composing the earth's crust oxygen forms one-half of their mass, and silicon another quarter, we find that in the meteorites these elements, though present, play a much less important part. The most abundant element in the meteorites is iron; and nickel, chromium, cobalt, manganese, sulphur, and phosphorus, are much more abundant in these extra-terrestrial bodies than they are in the earth's crust.

We have already referred to the remarkable fact that in our earth's crust nearly all the other elementary substances are found combined in the first instance with oxygen, and that most rocks consist of the oxide of silicon combined with the oxides of various metals. But this is by no means the case with the meteorites. In them we find metals like iron, nickel, cobalt, &c., in their uncombined condition, and forming alloys with one another. The same and other metals also occur in combination with carbon, phosphorus, chlorine, and sulphur, and some of the substances thus formed are quite unknown among terrestrial rocks. Compounds of the oxide of silicon with the oxides of the metals such as form the mass of the crust of the globe do occur in meteorites, but they play a much less important part than in the case of the terrestrial rocks.

Among the substances found in meteorites are several which do not exist among the terrestrial rocks—some, indeed, which it seems impossible to conceive of as being formed and preserved under terrestrial conditions. Among these we may mention the phosphide of iron and nickel (Schreibersite), the sulphide of chromium and iron (Daubréelite), the protosulphide of iron (Troilite), the sulphide of calcium (Oldhamite), the protochloride of iron (Lawrencite), and a peculiar form of crystallised silica, called by Professor Maskelyne 'Asmanite.'

There are other phenomena exhibited by meteorites which indicate that they must have been formed under conditions very different to those which prevail upon the earth's surface. Thus we find that fused iron and molten slag-like materials have remained entangled with each other, and have not separated as they would do if a great body like the earth were near to exercise the varying force of gravity upon the two classes of substances. Again, meteorites are found to have absorbed many times their bulk of hydrogen gas, and to exhibit peculiarities in their microscopic structure which can probably be only accounted for when we remember that they were formed in the interplanetary spaces, far away from any great attracting body.

But in recent years a number of very important facts have been discovered which may well lead us to devote a closer attention to the composition and structure of meteorites. It has been shown, on the one hand, that some meteorites contain substances precisely similar to those which are sometimes brought from the earth's interior during volcanic outbursts; and, on the other hand, there have been detected, among some of the ejections of volcanoes, bodies which so closely resemble meteorites that they were long mistaken for them. Both kinds of observation seem to point to the conclusion that the earth's interior is composed of similar materials to those which we find in the small planets called meteorites.

M. Daubrée has proposed a very convenient classification for meteorites, dividing them into the following four groups:—

I. Holosiderites; consisting almost entirely of metallic iron, or of iron alloyed with nickel, stony matter being absent; but sulphides, phosphides, and carbides of several metals are often diffused through the mass. The polished surfaces of these meteoric irons, when etched with acid, often exhibit a remarkable crystalline structure.

II. Syssiderites; in which a network of metallic iron encloses a number of granular masses of stony materials.

III. Sporadosiderites; which consist of a mass of stony materials, through which particles of metallic iron are disseminated.

IV. Asiderites; containing no metallic iron, but consisting entirely of stony materials.

There are, besides the meteorites belonging to these principal groups, a few of peculiar and exceptional composition, which we need not notice further for our present purpose.

From the above classification it will be seen that most meteorites consist of a mixture in varying proportions of metallic and stony materials. Sometimes the metallic constituents are present in greater proportions than the stony, at other times the stony materials predominate, while occasionally one or other of these elements may be wholly wanting.

The stony portions of meteorites, upon careful examination, prove to be built up of certain minerals, agreeing in their chemical composition and their crystalline forms with those which occur in the rocks of the earth's crust. Among the ordinary terrestrial minerals occurring in the stony portions of meteorites, we may especially mention olivine, enstatite, augite, anorthite, chromite, magnetite, and pyrrhotite.

The minerals which occur in meteorites are in every case such as are found in the more basic volcanic rocks—quartz, and the acid felspars, with the other minerals which occur in acid rocks, being entirely absent in the 'extra-terrestrial' rocks.

Now, besides the three great classes of lavas which we have described as being ejected from volcanic vents, there are some rarer materials occasionally brought from the earth's interior by the same agency, that present a most wonderful resemblance to the stony portions of meteorites. These materials we may call 'ultra-basic rocks.' Their specific gravity is very high, usually exceeding 3, and they contain a very low percentage of silica; on the other hand, the proportion of iron and magnesia is often much greater than in ordinary terrestrial rocks. But the most remarkable fact about these ultra-basic rocks is, that they are almost entirely composed of the minerals which occur in meteorites; namely, olivine, enstatite, augite, anorthite, magnetite, and chromite.

The ultra-basic rocks often occur under very peculiar conditions. Sometimes they are found forming ordinary volcanic protrusions through the sedimentary rocks. The rocks named pikrites, lherzolites, dunites, &c., are examples of such igneous protrusions composed of these ultra-basic materials, and probably all the true serpentines are rocks of the same class which have absorbed water and undergone great alteration. The ultra-basic rocks sometimes contain platinum and other metals in the free or uncombined state. But not unfrequently we find among the ordinary ejections of volcanoes, nodules and fragments of such ultra-basic materials, which have clearly been carried up with the other lavas from great depths in the earth's crust. Thus in Auvergne, the Eifel, Bohemia, Styria, and many other volcanic districts, the basaltic lavas and tuffs are found to contain nodules composed of the minerals which are so highly characteristic of meteorites. Such nodules, too, often form the centres of the volcanic bombs which are thrown out of craters during eruptions.

We thus see that materials identical in composition and character with the stony portions of meteorites, exist within the earth's interior, and are thrown out on its surface by volcanic action. A still more interesting discovery has been made in recent years; namely, that materials similar to the metallic portion of meteorites, and consisting of nickeliferous iron, also occur in deep-seated portions of the earth's crust, and are brought to the surface during periods of igneous activity.

In the year 1870, Professor Nordenskiöld made a most important discovery at Ovifak, on the south side of the Island of Disko, off the Greenland coast. On the shore of the island a number of blocks of iron were seen, and the chemical examination of these proved that, like ordinary metallic meteorites, they consisted of iron alloyed with nickel and cobalt.

Now, when the facts concerning the masses of native iron of Ovifak were made known, the first and most natural explanation which presented itself to every mind was, that these were a number of meteorites which at some past period had fallen upon the earth's surface.

But a further examination of the locality revealed a number of facts which, as Professor Steenstrup pointed out, it is very difficult to reconcile with the theory that the Ovifak masses of iron are of meteoric origin. The district of Western Greenland, where these masses were discovered, has been the scene of volcanic outbursts on the grandest scale during the Miocene period. In close proximity to the great iron masses, there are seen a number of basaltic dykes; and, when these dykes are carefully examined, the basaltic rock of which they are composed is seen to be full of particles of metallic iron. In fig. 87, we have a drawing made from a section of the Ovifak basalts magnified four or five diameters. The rock-mass is seen to be composed of black, opaque magnetite, and transparent crystals of augite, labradorite, olivine, &c.; while, through the whole, particles of metallic iron are found entangled among the different crystals in the most remarkable manner.

It has been suggested that this singular rock might have been formed by a meteorite falling, in Miocene times, into a lava-stream in a state of incandescence. But the relation of the metallic particles to the stony materials is such as to lend no support whatever to this rather strained hypothesis.

A careful study of all the facts of the case by Lawrence Smith, Daubrée, and others well acquainted with the phenomena exhibited by meteorites, has led to the conclusion that the large iron-masses of Ovifak, as well as the particles of metallic iron diffused through the surrounding basalts, are all of terrestrial origin, and have been brought by volcanic action from the earth's interior. It is probable that, just as we find in many basaltic lavas nodules of ultra-basic materials similar to the stony parts of meteorites, so in these basalts of Ovifak we have masses of iron alloyed with nickel, similar to the metallic portions of meteorites. Both the stony and metallic enclosures in the basalt are in all probability derived from deeper portions of the earth's crust. By the weathering away of the basalt of Ovifak, the larger masses of metallic iron have been left exposed upon the shore where they were found.

There are a number of other facts which seem to support this startling conclusion. Thus it has been shown by Professor Andrews that certain basalts in our own islands contain particles of metallic iron of microscopic dimensions, and it is not improbable that some of the masses of nickeliferous iron found in various parts of the earth's surface, which have hitherto been regarded as meteorites, are, like those of Ovifak, of terrestrial origin.

Another piece of evidence pointing in the same direction, is derived from those great fissures communicating with the interior of our globe which become filled with metallic minerals, and are known to us as mineral-veins. In these mineral-veins the native metals, their alloys, and combinations of these with sulphur, chlorine, phosphorus, &c., are frequently present. But oxides of the metals, except as products of subsequent alteration, occur far less frequently than in the earth's crust generally. Hence we are led to conclude that the substances which in the outer part of the earth's crust always exist in combination with oxygen, are at greater depths in a free and uncombined condition.

Nor is it a circumstance altogether unworthy of attention that the researches of Mr. Norman Lockyer and other astronomers, based on the known facts of the relative densities of the several members of the solar system, and the ascertained relations of the different solar envelopes, have led to conclusions closely in accord with those arrived at by geologists. These researches appear to warrant the hypothesis that the interior of our globe consists of metallic substances uncombined with oxygen, and that among these metallic substances iron plays an important part. Our globe, as we know, is a great magnet, and the remarkable phenomena of terrestrial magnetism may also not improbably find their explanation in the fact that metallic iron forms 80 large a portion of the earth's interior.

The interesting facts which we have been considering may be made clearer by the accompanying diagram (fig. 88). The materials ejected from volcanic vents (lavas) are in almost all cases compounds of silicon and the various metals with oxygen. In the lighter or acid lavas oxygen constitutes one-half of their weight, and the proportion of metals of the iron-group is very small. As we pass to the heavier intermediate and basic lavas, we find the proportion of oxygen diminishing, and the metals of the alkaline earths (magnesium and calcium) with the metals of the iron-group increasing, in quantity. In the small and interesting group of the ultra-basic lavas the proportion of oxygen is comparatively small, and the proportion of magnesium and iron very high. So much for the terrestrial rocks.

Now let us turn our attention to the extra-terrestrial rocks or those found in meteorites. The Asiderites are quite identical in composition with the ultra-basic lavas of our globe, but in the Sporadosiderites and the Syssiderites we find the proportion of oxygen rapidly diminishing, and that of metallic iron increasing. Finally, in the Holosiderites the oxygen entirely disappears, and the whole mass becomes metallic.

From the Holosiderites at one end of the chain to the add lavas at the other, we find there is a complete and continuous series; the rocks of terrestrial origin overlapping, in their least oxydized representatives, the most highly oxydized representatives of the extra-terrestrial rocks. But the discovery at Ovifak of the iron-masses, and the basalts with iron disseminated, has afforded another very important link, placing the terrestrial and extra-terrestrial rocks in closer relations with one another.

All these facts appear to point to the conclusion that the earth's interior consists of metallic substances either quite uncombined or simply alloyed with one another, and among these iron is very conspicuous by its abundance. The outer crust, which is probably of no great thickness, contains an enormous proportion of oxygen and silicon combined with the materials which constitute the interior portions of our globe. It may be, as has been suggested by astronomers, that our earth consisted at one time of a solid metallic mass surrounded by a vaporous envelope of metalloids, and that the whole of the latter, with the exception of the constituents of the atmosphere and ocean, have gradually entered into combination with the metals of the nucleus to form the existing crust of the globe. But of this period the geologist can take no cognisance. The records which he studies evidently commenced at a long subsequent period, when the conditions prevailing at the earth's surface differed but little, if at all, from those which exist at the present day. Equally little has the geologist to do with speculations concerning a far distant future when, as some philosophers have suggested, the work of combination of the waters and atmosphere of the earth's surface with the metallic substances of its interior shall be completed, and our globe, entirely deprived of its fluid envelopes, reduced to the condition in which we find our satellite, the moon.

There is another class of enquiries concerning the earth's interior to which the attention of both geologists and astronomers has long been directed—that, namely, which deals with the problem of the physical condition of the interior of our globe.

The fact that masses of molten materials are seen at many points of the earth's surface to issue from figures in the crust of our globe, seems at first sight to find a simple explanation if we suppose our planet to consist of a fluid central mass surrounded by a solid crust. Hence we find that among those who first thought upon this subject, this hypothesis of a liquid centre and a solid crust was almost universally accepted. This hypothesis was supposed to find further support in the fact that, as we penetrate into the earth's crust by mines or boring operations, the temperature is found to continually increase. It was imagined, too, that this condition of our planet would best agree with the requirements of the nebular hypothesis of Laplace, which explains the formations and movements of the bodies of the solar system by the cooling down of a nebulous mass.

But a more careful and critical examination of the question has led many geologists and astronomers to reject the hypothesis that the earth consists of a great fluid mass surrounded by a comparatively thin shell of solid materials.

Volcanic outbursts and earthquake tremors, though so terrible and destructive to man and his works, are but slight and inconsiderable disturbances in a globe of such vast dimensions as that on which we live. The condition of the crust of the globe is, in spite of volcanic and earthquake manifestations, one of general stability; and this general stability has certainly been maintained during the vast periods covered by the geological record. Such a state of things seems quite irreconcilable with the supposition that, at no great depth from the surface, the whole mass of the globe is in a liquid condition. If, on the other hand, it be supposed that the solid crust of the globe is several hundreds of miles in thickness, it is difficult to understand how the local centres of volcanic activity could be supplied from such deep-seated sources.

There are other facts which seem equally irreconcilable with the hypothesis of a fluid centre and a thin solid crust in our globe. If all igneous products were derived from one central reservoir, we might fairly expect to find a much greater uniformity of character among those products than really exists. But in some cases, materials of totally different composition are ejected at the same time from closely adjoining volcanic districts. Thus in Hungary and Bohemia, as we have seen, lavas of totally different character were being extruded during the Miocene period. In the island of Hawaii, as Professor Dana has pointed out, igneous ejections have taken place at a crater 14,000 feet above the sea-level, while a closely adjoining open vent at a level 10,000 feet lower exhibited no kind of sympathy with the disturbance. Whatever may be the cause of volcanic action, it seems clear that it does not originate in a universal mass of liquefied material situated at no great depth from the earth's surface.

The conclusions arrived at by astronomers and physicists is one quite in accord with those which geologists have reached by totally different methods. It is now very generally admitted that if the earth were not a rigid mass, its behaviour under the attract live influences of the surrounding members of the solar system would be very different to what is found to be the case.

That the earth is in a solid condition to a great depth from the surface, and possibly quite to the centre, is a conclusion concerning which there can be little doubt; and in the next chapter we shall endeavour to show that such a condition of thirds is by no means incompatible with those manifestations of internal energy, the phenomena of which we are considering in this work. The question, therefore, of the complete solidity of our globe, or of its consisting of a solid and a liquid portion, is one of speculative interest only, and is in no way involved in our investigations concerning the nature and origin of volcanic activity. We may conclude this chapter by enumerating the several hypotheses which have at different times been maintained concerning the nature of the interior of our globe.

First. It has been suggested that the earth consists of a fluid or semi-fluid nucleus surrounded and enclosed in a solid shell. Some have maintained this shell to be of such insignificant thickness, as compared with the bulk of the interior liquid mass, that portions of the latter are able to reach the earth's surface through movements and fractures of the outer shell, and that in this manner volcanic manifestations originate. Others, impressed with the general stability and rigidity of the globe as a whole, have maintained that the outer solid shell must have a very considerable thickness, amounting probably to not less than several hundreds of miles. But through a shell of such thickness it is difficult to conceive of the liquid masses of the interior finding their way to the surface, and those who have held this view are driven to suggest some other means by which local developments of volcanic action might be brought about.

Secondly. Some physicists have asserted that a globe of liquid matter radiating its heat into space, would tend to solidify both at the surface and the centre, at the same time. The consequence of this action would be the production of a sphere with a solid external shell and a solid central nucleus, but with an interposed layer in a fluid or semi-fluid condition. It has been pointed out that if we suppose the solidification to have gone so far, as to have caused the partial union of the interior nucleus and the external shell, we may conceive a condition of things in which the stability and rigidity is sufficient to satisfy both geologists and astronomers, but that in still unsolidified pockets or reservoirs, filled with liquefied rock, between the nucleus and the shell, we should have a competent cause for the production of the volcanic phenomena of the globe. In this hypothesis, however, it is assumed that the cooling at the centre and the surface of the globe would go on at such rates that the reservoirs of liquid material would be left at a moderate depth from the surface, so that easy communication could be opened between them and volcanic vents.

Thirdly. It has been maintained that the earth may have become perfectly solid from the centre to the surface. Those who hold this view endeavour to account for the phenomena of volcanoes in one of two ways. It may be, they say, that the deep-seated rock-masses, though actually solid, are in a state of potential liquidity; that though reduced to a solid state by the intense pressure of the superincumbent masses, yet such is the condition of unstable equilibrium in the whole mass, that the comparatively slight movements and changes taking place at the earth's surface suffice to bring about the liquefaction of portions of its crust and consequent manifestations of volcanic energy. But It may be, as other supporters of the doctrine of the earth's complete solidity have maintained, that the phenomena of volcanoes have no direct connection with a supposed incandescent condition of our planet at all, and that there are chemical and mechanical forces at work within our globe which are quite competent to produce at the surface all those remarkable phenomena which we identify with volcanic action.

From this summary of the speculative views which have been entertained upon the subject of the physical condition of the earth's interior, it will be clear that at present we have not sufficient evidence for arriving at anything like a definite solution of the problem. The conditions of temperature and pressure which exist in the interior of a globe of such vast dimensions as our earth, are so far removed from those which we can imitate in our experimental enquiries, and it is so unsafe to push the application of laws arrived at by the latter to the extreme limits required by the former, that we shall do well to pause before attempting to dogmatise on such a difficult question.

In the next chapter we shall endeavour to grapple with a somewhat more hopeful task, to point out how far observation and experiment have enabled us to offer a reasonable explanation of the wonderful series of phenomena which are displayed during outbursts of volcanic activity.